Post
JA EN

The Science of Age and Cognitive Abilities—What Declines and What Grows (Part 1/3)

Posted
By
16 min read
The Science of Age and Cognitive Abilities—What Declines and What Grows (Part 1/3)
  • Target Audience: Software engineers, IT professionals interested in AI adoption
  • Prerequisites: None
  • Reading Time: 15 minutes
  • Series: Part 1 of 3

About This Series

Research shows that “language ability peaks in your 50s.” So what should you acquire by age 50 for the AI era?

This series (3 parts total) reviews the relationship between age and cognitive abilities based on cognitive science research, redefines the value of experience in the AI era, and presents specific actions to take before cognitive decline.

Series Structure:

Overview

“You get dumber as you age”—this naive perception is half right and half wrong.

What cognitive science research has revealed is that cognitive abilities don’t decline along a single curve, but follow different trajectories for different abilities. Processing speed peaks in the late teens, and working memory peaks around age 30. Meanwhile, vocabulary continues to improve into the 50s and 60s, and emotion recognition ability reaches its peak between ages 40 and 50.

This article organizes scientific findings on age and cognitive abilities, centering on the large-scale research by Hartshorne & Germine (2015). This isn’t easy encouragement that “you’ll be fine as you age.” By accurately understanding what is lost and what is gained, we provide a foundation for strategically designing your career.

Cognitive Abilities Don’t Peak at One Age

Hartshorne & Germine (2015) Discovery

In 2015, MIT and Harvard researchers Joshua Hartshorne and Laura Germine analyzed approximately 48,500 web participants and normative data from standardized IQ and memory tests, publishing groundbreaking research on peak ages for cognitive abilities1.

“Our results reveal considerable heterogeneity in when cognitive abilities peak. Some abilities peak and begin to decline around the time of high school graduation; others plateau in early adulthood and begin to decline in the 30s; still others do not peak until people reach their 40s or later.”1

This discovery overturned the naive perception that “you’re smartest when you’re young.”

Peak Ages by Ability

flowchart TB
    subgraph Peak["Peak Ages for Cognitive Abilities"]
        direction TB
        A["Processing Speed<br>(18-19 years)"]
        B["Working Memory<br>(~30 years)"]
        C["Face Recognition<br>(~32 years)"]
        D["Emotion Recognition<br>(40-60 years)"]
        E["Vocabulary<br>(50-65 years)"]
    end

    A --> B
    B --> C
    C --> D
    D --> E

Key peak ages for cognitive abilities shown by Hartshorne & Germine’s research:

Cognitive AbilityPeak AgeCharacteristics
Processing Speed18-19 yearsEarliest peak, followed by linear decline
Short-term Memory~25 yearsRelatively early peak
Working Memory~30 yearsBoth verbal and visual peak around same time
Face Recognition~32 yearsPeaks later than processing speed
Emotion Recognition40-60 yearsWide peak period, gradual decline after 60
Vocabulary50-65 yearsLatest peak, further delayed by generational effects in recent years

“At any given age, individuals improve on some abilities, decline on others, and plateau on still others. There probably is no age at which humans are at their peak on most abilities.”1

Vocabulary Peaks Even Later Now

Interestingly, the peak age for vocabulary has been changing across generations2:

  • 1974-1987: Peak in early 40s
  • 1988-1997: Peak around age 50
  • 1998-2012: Peak at age 65

Researchers attribute this change to the following factors:

  • Rising education levels
  • Increase in white-collar jobs requiring reading and writing
  • Improved nutrition and physical health
  • Increased intellectual stimulation opportunities for older adults

Fluid Intelligence and Crystallized Intelligence

Two Types of Intelligence

Cognitive psychology broadly classifies intelligence into two categories3:

Fluid Intelligence (Gf)

  • Ability to solve new problems
  • Abstract reasoning, pattern recognition
  • Processing speed, working memory
  • Peaks around age 20, then declines

Crystallized Intelligence (Gc)

  • Accumulated knowledge and experience
  • Vocabulary, general knowledge, specialized skills
  • Continues to improve until 60s-70s
flowchart TB
    subgraph Fluid["Fluid Intelligence (Gf)"]
        direction LR
        F1["Processing Speed"]
        F2["Working Memory"]
        F3["Abstract Reasoning"]
        F4["Pattern Recognition"]
    end

    Fluid --> Peak1["Peak: Around age 20"]

    subgraph Crystallized["Crystallized Intelligence (Gc)"]
        direction LR
        C1["Vocabulary"]
        C2["General Knowledge"]
        C3["Specialized Skills"]
        C4["Procedural Knowledge"]
    end

    Crystallized --> Peak2["Peak: 60s-70s"]

Mechanism of Fluid Intelligence Decline

Salthouse (1996)’s “Processing Speed Theory” is a representative theory explaining the mechanism of age-related cognitive decline4:

“Adult age is associated with a decrease in the speed at which many processing operations can be executed, and this speed reduction leads to impairments in cognitive functioning through ‘limited time’ mechanisms and ‘simultaneity’ mechanisms.”4

Limited Time Mechanism When processing is slow, necessary operations cannot be completed within the time limit.

Simultaneity Mechanism When processing is slow, early processing results are lost before later processing is completed.

This theory has been supported by many empirical studies, showing that processing speed decline is a major factor in other cognitive declines such as working memory and executive function5.

Why Crystallized Intelligence Is Maintained and Improves

On the other hand, research points to the following factors for why crystallized intelligence is maintained and improves with age6:

  1. Stability of Semantic Memory: The semantic memory system that stores vocabulary and factual knowledge is less affected by age
  2. Cumulative Effect: Knowledge accumulates over time and facilitates new learning
  3. Diversity of Neural Circuits: Crystallized intelligence is distributed across multiple brain regions, making it resistant to damage in any single area

“Semantic memory scores rose from 35 to 55 years of age, maintained stability until 65, then showed a small decrease after 65. The researchers concluded that semantic representations are maintained in older adults.”6

What Declines and What Grows

Abilities That Definitely Decline

Cognitive abilities that research consistently shows decline with age:

1. Processing Speed45

  • Declines about 0.02 standard deviations per year after age 20
  • In the 60s, about 60-70% of the speed in the 20s
  • Broadly affects everything from simple reaction time to complex cognitive tasks

2. Working Memory7

  • Gradual decline after age 30
  • Particularly noticeable in complex tasks (manipulating multiple pieces of information simultaneously)
  • Decline is slower with higher education levels

3. Episodic Memory8

  • Memory of specific events and experiences
  • Details of “when, where, and what happened”
  • Recall accuracy declines after age 40

4. Divided Attention9

  • Ability to process multiple tasks simultaneously
  • Decreased multitasking performance

Abilities That Are Maintained or Improve

1. Vocabulary and Language Knowledge16

  • Continues to improve until 50s-60s
  • Retrieval takes longer, but knowledge itself is rich
  • “Tip-of-the-tongue” phenomenon increases, but knowledge is preserved

2. Semantic Memory6

  • Facts, concepts, word meanings
  • Hardly affected by age
  • Rather, becomes richer through experience

3. Emotion Recognition and Social Cognition1

  • Peaks in 40s-50s
  • Ability to read others’ emotions
  • Only gradual decline after age 60

4. Specialized Knowledge and Procedural Memory10

  • “How to do” knowledge is maintained
  • Mastered skills are less affected by age
  • Automated behavioral patterns

Compensation Through Experience

Air Traffic Controller Study

Decline in cognitive function doesn’t necessarily lead to decline in job performance. Morrow et al. (2003)’s study of air traffic controllers clearly demonstrates the compensation mechanism through experience10:

“On simple cognitive tasks—unrelated to their professional lives as controllers—older controllers were slower than their younger colleagues. But when it came to job-related tasks, the results were roughly even.”

Study Details:

  • Compared young controllers (20-27 years), older controllers (53-64 years), and age-matched non-controllers
  • Basic cognitive tests: Older controllers slower than young
  • Job simulations: No difference between older and young controllers
  • Older non-controllers: Clear performance decline

These results show that experience in a specialized domain can compensate for general cognitive function decline.

Mechanisms of Compensation

Researchers have proposed three mechanisms for why experts can maintain performance despite cognitive decline11:

1. Preserved Differentiation Basic abilities in the specialized domain are maintained better than general cognitive functions.

2. Compensation Declining abilities are supplemented by other abilities or strategies. For example, typists may maintain overall speed by increasing lookahead even when finger movements slow.

3. Selective Maintenance Limited cognitive resources are concentrated in the most important domains.

Job Performance and Age

Multiple meta-analyses show important findings about the relationship between age and job performance12:

“Little relationship was shown between age and job performance. Standardized cognitive tests do not capture the complexity of work situations, and in many job tasks, older workers can draw on knowledge, experience, and contextual support mechanisms to compensate for age-related changes in cognitive abilities.”12

Changes in Cognitive Flexibility and Learning Ability

Compensation through experience is powerful, but there is one important constraint. Cognitive flexibility—the ability to switch thinking and adapt to new information—declines with age.

Decline in Cognitive Flexibility

A 2024 meta-analysis clearly shows the relationship between cognitive flexibility and age13:

“Groups over 60 years of age have significantly lower cognitive flexibility compared to young adults aged 18-35.”

The decline in cognitive flexibility is associated with reduced gray matter volume in the frontal lobe, changes in white matter integrity, and decreased dopamine system function.

Proactive Interference—Old Knowledge Hinders New Learning

Another important phenomenon with aging is decreased ability to cope with Proactive Interference14:

“Higher age is associated with reduced interference control, and age-related cognitive decline is linked to impaired interference control.”

Proactive interference is a phenomenon where previously learned information interferes with memory and retrieval of new information. For example, programming patterns used for many years may hinder acquisition of new paradigms. This is a situation where “experience gets in the way,” showing that “experience is good” isn’t simply true.

Advantage of Learning That Leverages Existing Knowledge

However, not all learning becomes difficult for older adults. Learning that relates to existing knowledge shows that older adults achieve equal or better outcomes than young people15:

“Older adults have declining episodic memory, but semantic memory is maintained. Memory for information consistent with existing schemas (knowledge frameworks) is more effective in older adults.”

The National Academies report also includes cases where older adults learned more than young people in fields where they had existing knowledge, such as health and finance learning research.

Implications for Age-Specific Learning Strategies

These findings suggest that optimal learning strategies differ by age:

AgeCognitive CharacteristicsRecommended Learning Strategy
40sCognitive flexibility still relatively highUnlearning: Good opportunity to review old habits
50s and beyondIncreasing proactive interferenceLeveraging existing knowledge: Integrate new information with existing frameworks

The 40s can be called the “last good opportunity for unlearning.” During this period when cognitive flexibility is still maintained, consciously reviewing outdated knowledge and habits determines adaptability in the 50s and beyond.

Detailed learning strategies will be explained in Part 3, “Action Plan to Start Now.”

Implications for the AI Era—50s Are the “Right Age” for AI Utilization

Based on the scientific findings so far, one conclusion emerges. Your 50s may be the most advantageous age for AI utilization.

AI Compensates for the Limits of Fluid Intelligence

In the AI era, decline in processing speed and working memory becomes less of a problem in a sense:

  • Information Processing: AI instantly processes large amounts of data
  • Memory: AI functions as external memory
  • Parallel Processing: AI agents execute multiple tasks simultaneously

The abilities that definitely decline in your 50s—processing speed, working memory, divided attention—are precisely the areas where AI excels.

The Value of Crystallized Intelligence Increases

On the other hand, crystallized intelligence is essential for effectively utilizing AI:

  • Precision of AI Instructions: Rich vocabulary and expressive ability enable accurate prompt creation
  • AI Output Evaluation: Without deep specialized knowledge, you can’t judge the quality of AI output
  • Context Understanding: Tacit knowledge based on experience enables appropriate context setting

“What makes experts ‘high quality’ is not that they use AI efficiently, but that they rigorously scrutinize AI output”—From the existing article “The Truth About Experts Who Appear to ‘Delegate Everything to AI’”

Why Language Ability in Your 50s Maximizes AI Utilization

The fact that vocabulary peaks between ages 50-65 has decisive significance in the AI era.

AI is a tool “operated through language.”

Just as programming operates machines through code, AI is operated through prompts (natural language). And prompt precision is directly linked to language ability.

flowchart LR
    subgraph Fifties["Cognitive Profile in 50s"]
        direction TB
        V["Vocabulary: Peak"]
        S["Specialized Knowledge: Accumulated"]
        E["Emotion Recognition: Peak Period"]
        M["Metacognition: Mature"]
    end

    subgraph AI_Use["Fit for AI Utilization"]
        direction TB
        P["Precise Prompt Creation"]
        J["Quality Judgment of Output"]
        C["Team Collaboration"]
        R["Recognition of Own Limits"]
    end

    V --> P
    S --> J
    E --> C
    M --> R

Specific reasons why 50s have advantages in AI utilization:

  1. Can convey subtle nuances: Rich vocabulary enables instructions of “exactly this” rather than “something like this”

  2. Can eliminate ambiguity: From years of experience, you know what’s ambiguous and what’s clear. You can appropriately select what context to convey to AI

  3. Can detect AI’s “plausible mistakes”: Accumulated specialized knowledge allows detection of errors AI outputs with confidence

  4. Can give instructions at appropriate abstraction level: You understand the “just right granularity” of instructions that are neither too detailed nor too vague

Paradigm Shift from “Speed” to “Precision”

When young engineers use AI, their high processing speed enables an approach of “trying many things and selecting what works.”

Engineers in their 50s can take a different approach. Give high-precision instructions from the start and achieve high-quality results with fewer attempts.

This isn’t “slow.” It’s eliminating wasteful attempts.

ApproachCharacteristicsSuitable Age
High-Speed Trial TypeTry many prompts and select good resultsYoung people with high processing speed
Precision Instruction TypeGive accurate instructions from start, get results with few attempts50s with peak language ability

Neither is “superior.” What’s important is choosing an approach that matches your cognitive profile.

Next Time

This article organized scientific findings on age and cognitive abilities and showed why those in their 50s have advantages in AI utilization. However, there’s an important premise for experience to have value—it must be “experience that keeps being updated.” In Part 2, we’ll delve deeper into the “value of experience” in the AI era, exploring the true meaning of research results showing experts “slow down” when using AI, and the “shelf life” of experience.

Summary

  1. Cognitive abilities don’t decline along a single curve: Different abilities follow different trajectories, and vocabulary and emotion recognition don’t peak until after the 50s
  2. Fluid intelligence declines early: Processing speed peaks around age 20, working memory around age 30
  3. Crystallized intelligence continues to improve: Vocabulary, semantic memory, and specialized knowledge grow until the 60s-70s
  4. Experience compensates for cognitive decline: In specialized domains, experience compensates for processing speed decline
  5. Cognitive flexibility declines: Ability to switch thinking and adapt to new learning declines with age, and the 40s are the “last good opportunity for unlearning”
  6. The AI era increases the value of crystallized intelligence: AI complements fluid intelligence, relatively increasing the value of experience and knowledge

See also other articles related to this theme:

References

References corresponding to citation numbers in the text are listed in numerical order.

Additional References (Not Numbered in Text)

About Citation Accuracy: The research cited in this article has been verified through the following methods:

  • Confirmation via academic databases (PubMed, Google Scholar)
  • Verification of paper information on official journal websites
  • Cross-verification through multiple independent sources (academic media, official announcements from research institutions, etc.)

  1. When Does Cognitive Functioning Peak? The Asynchronous Rise and Fall of Different Cognitive Abilities Across the Life Span - Hartshorne, J. K., & Germine, L. T. (2015). Psychological Science, 26(4), 433-443. 【Reliability: High】 ↩︎ ↩︎2 ↩︎3 ↩︎4 ↩︎5

  2. The rise and fall of cognitive skills - MIT News (2015). 【Reliability: Medium-High】 ↩︎

  3. A strong dependency between changes in fluid and crystallized abilities in human cognitive aging - Buschke, H., et al. (2022). Scientific Reports. 【Reliability: High】 ↩︎

  4. The processing-speed theory of adult age differences in cognition - Salthouse, T. A. (1996). Psychological Review, 103(3), 403-428. 【Reliability: High】 ↩︎ ↩︎2 ↩︎3

  5. The Impact of Age on Cognition - Murman, D. L. (2015). Seminars in Hearing, 36(3), 111-121. 【Reliability: High】 ↩︎ ↩︎2

  6. Semantic memory, but not education or intelligence, moderates cognitive aging - Roquet, D., et al. (2020). Neurobiology of Aging, 85, 104-114. 【Reliability: High】 ↩︎ ↩︎2 ↩︎3 ↩︎4

  7. Working memory in older adults declines with age, but is modulated by sex and education - Pliatsikas, C., et al. (2019). Quarterly Journal of Experimental Psychology, 72(6), 1308-1327. 【Reliability: High】 ↩︎

  8. Episodic and semantic memory functioning in very old age - Nyberg, L., et al. (2015). Cogent Psychology. 【Reliability: High】 ↩︎

  9. Characterizing and Assessing Cognitive Aging - National Academies of Sciences (2015). Cognitive Aging: Progress in Understanding and Opportunities for Action. 【Reliability: High】 ↩︎

  10. Experience-Based Mitigation of Age-Related Performance Declines: Evidence From Air Traffic Control - Morrow, D. G., et al. (2003). Journal of Experimental Psychology: Applied, 9(2), 94-106. 【Reliability: High】 ↩︎ ↩︎2

  11. Aging and Expertise - Krampe, R. T., & Charness, N. (2006). In K. A. Ericsson, N. Charness, P. J. Feltovich, & R. R. Hoffman (Eds.), The Cambridge Handbook of Expertise and Expert Performance (pp. 723-742). Cambridge University Press. 【Reliability: High】 ↩︎

  12. Consequences of Age-Related Cognitive Declines - Salthouse, T. A. (2012). Annual Review of Psychology, 63, 201-226. 【Reliability: High】 ↩︎ ↩︎2

  13. A meta‐analysis of cognitive flexibility in aging: Perspective from functional network and lateralization - Human Brain Mapping (2024). 【Reliability: High】 ↩︎

  14. Proactive interference in working memory is related to adult age and cognitive factors - Samrani, G., & Persson, J. (2021). Aging, Neuropsychology, and Cognition, 28(1), 58-76. 【Reliability: High】 ↩︎

  15. Learning Across the Life Span - National Academies of Sciences (2018). How People Learn II: Learners, Contexts, and Cultures. 【Reliability: High】 ↩︎

AI・Technology
Cognitive-Science Aging Career AI Expertise
This post is licensed under CC BY 4.0 by the author.